You can only download them using a free ACM account, and even with such account, you can't download most of them (ie, von Neumann's one).
I was going to ask about that - the Neumann one was the only one I found interesting, but I could not find an actual download link.
The ACM is so utterly backward that I won't even click on their links anymore. In an era when nearly everything else is a 5-second web search away, they place barrier after barrier in front of information. What are they protecting other than their own obsolescence? Googling the titles of most papers brings up a free copy on the author's website or elsewhere.
Agreed that the ACM's Digital Library is a total waste of a person's time and always has been. Practically every important paper in physics and astronomy is available from a central location (arxiv.org) without restriction and with two clicks (one to specify the paper, the second to choose between PDF, Postscript, Latex source code, etc), and it is only a matter of time before computer science is the same way.